[Nek5000-users] Change boundary condition type with time?

[nek5000-users at lists.mcs.anl.gov]

Dear Johan,


I guess I'm confused about the "open boundary" part of your request.


Do you anticipate flow leaving through this part? Or no flow coming in?

The reason I ask is that it's generally unstable to have any flow coming into

the domain through an open boundary.


Aside from that, there are some technical issues to deal with.  You need to

change the boundary conditions, as Nicolas pointed out, but you also need to

recompute the preconditioners that are set up at the beginning of the computation

because they depend on the type of BC that is employed.


I think the easiest and best approach would be to evolve your initial condition to

the point where you want to change the BC and then stop the calculation and

use the second BC option from that point (i.e., by using a different rea file for

the second part of the computation, or by applying Nicolas's suggestion in

subroutine usrdat2, which is called before the preconditioners are established).

Note that you can have your inflow BC be time dependent -- flow turned, then flow

turned off -- which is all Dirichlet, and you don't need to start/stop the calculation.


hth


Paul


________________________________
From: Nek5000-users <nek5000-users-bounces at lists.mcs.anl.gov> on behalf of nek5000-users at lists.mcs.anl.gov <nek5000-users at lists.mcs.anl.gov>
Sent: Wednesday, April 11, 2018 7:08:12 AM
To: nek5000-users at lists.mcs.anl.gov
Subject: [Nek5000-users] Change boundary condition type with time?


Hi Neks,


I wonder if it is possible to have boundary conditions that chage type with time?


I want to simulate a flow where I make a 'puff' of fluid through a hole in a wall.

After the puff is made it continues to evolve without beeing forced.


Mathematically, this is what I want:

  1.  Dirichlet boundary conditions (ux,uy,uz) = (0,0,1) for the velocity at a part B of the boundary up to a time T.
  2.  Open boundary condition for the velocity at B for time > T.


Is this possible to implement?


Best,


Johan

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/nek5000-users/attachments/20180412/a44f5a7a/attachment.html>